2026-09论文中文摘要
ReLoc：重新思考用于鲁棒室外基于LiDAR定位的场景坐标回归架构
ReLoc: Rethinking Scene Coordinate Regression Architecture for Robust Outdoor LiDAR-based Localization
arXiv 2609.33344 · 2026-09-29CHI：一种统一实体、关系和数量维度的复合幻觉指数，用于摘要评估
CHI: A Composite Hallucination Index Unifying Entity, Relation, and Quantity Dimensions for Summarization Evaluation
arXiv 2609.33343 · 2026-09-29CalibHyper：用于少样本分子性质预测的机会校正关系超图
CalibHyper: Chance-Corrected Relational Hypergraphs for Few-Shot Molecular Property Prediction
arXiv 2609.33342 · 2026-09-29离子温度梯度湍流从有限到弱磁剪切区域
Ion-temperature-gradient turbulence from finite to weak magnetic shear regime
arXiv 2609.33341 · 2026-09-29RAEGL：面向时间偏移下选择性上下文路由的风险感知证据门控学习
RAEGL: Risk-Aware Evidence-Gated Learning for Selective Contextual Routing under Temporal Shift
arXiv 2609.33340 · 2026-09-29自然度引导的流形流匹配用于手语生成
Naturalness-guided Manifold Flow Matching for Sign Language Production
arXiv 2609.33339 · 2026-09-29OPERA：用于指代视频分割的统一全模态渐进式时空推理智能体
OPERA: A Unified Omnimodal Progressive Spatio-Temporal Reasoning Agent for Referring Video Segmentation
arXiv 2609.33338 · 2026-09-29安全分数匹配：基于Hamilton-Jacobi可达性的扩散策略用于在线安全强化学习
Safe Score Matching: Diffusion Policies with Hamilton-Jacobi Reachability for Online Safe Reinforcement Learning
arXiv 2609.33337 · 2026-09-29超越保守：基于可恢复性条件的探索用于基于模型的模仿学习
Beyond Conservatism: Recoverability-Conditioned Exploration for Model-Based Imitation Learning
arXiv 2609.33336 · 2026-09-29学习预测世界是否有助于智能体行动？审计世界模型后训练
Does Learning to Predict the World Help Agents Act? Auditing World-Model Post-Training
arXiv 2609.33335 · 2026-09-29最大外平面图的最小谱半径
The minimum spectral radius of maximal outerplanar graphs
arXiv 2609.33333 · 2026-09-29CertMark：具有认证解码的无失真多比特水印
CertMark: Distortion-Free Multi-Bit Watermarking with Certified Decoding
arXiv 2609.33332 · 2026-09-29Conway-Parker 代数与最大 Fischer 群
The Conway-Parker algebra and the largest Fischer group
arXiv 2609.33331 · 2026-09-29FeCoSplat：用于前馈3D高斯泼溅的反馈引导压缩
FeCoSplat: Feedback-Guided Compression for Feed-Forward 3D Gaussian Splatting
arXiv 2609.33330 · 2026-09-29Morse-Bott不等式在Lefschetz复形上的应用
Morse-Bott inequalities on Lefschetz complexes
arXiv 2609.33329 · 2026-09-29在格子玻尔兹曼方法中，矩基底何时重要？
When does the moment basis matter in lattice Boltzmann methods?
arXiv 2609.33328 · 2026-09-29素数域上的平面代数Zarankiewicz定理
A planar algebraic Zarankiewicz theorem over prime fields
arXiv 2609.33327 · 2026-09-29ANTMAN：大规模信息空间中多智能体导航的自适应需求追踪
ANTMAN: Adaptive Need Tracking for Multi-Agent Navigation in Large Information Spaces
arXiv 2609.33326 · 2026-09-29VisionHOPE：视觉骨干网络作为自修改学习系统
VisionHOPE: Visual Backbones as Self-Modifying Learning Systems
arXiv 2609.33325 · 2026-09-29关于Euler双zeta函数$\zeta_2(s,s)$零点分布的研究
On the distribution of the zeros of the Euler double zeta-function $ζ_2(s,s)$
arXiv 2609.33324 · 2026-09-29智能体多轮推理：一种公平性方法
Agentic Multi-Turn Reasoning: A Fairness Approach
arXiv 2609.33323 · 2026-09-29面向智能体文档分析的鲁棒层次结构
Robust Hierarchical Structures for Agentic Document Analysis
arXiv 2609.33322 · 2026-09-29多模光源脉冲波动在具有任意色散的光纤和波导中的传播
Propagation of pulse fluctuations of multimode sources in optical fibers and waveguides with arbitrary chromatic dispersion
arXiv 2609.33320 · 2026-09-29PhysAlign：多模态物理推理中基于证据的角色对齐基准
PhysAlign: A Benchmark for Evidence-Grounded Role Alignment in Multimodal Physics Reasoning
arXiv 2609.33319 · 2026-09-29PIC-UIE：轻量级水下图像增强的图像自适应校正预测
PIC-UIE: Predicting Image-Adaptive Corrections for Lightweight Underwater Image Enhancement
arXiv 2609.33318 · 2026-09-29太阳活动周24期间D区电离层电子密度变化建模
Modeling of the variability of D-region ionospheric electron density during solar cycle-24
arXiv 2609.33317 · 2026-09-29Navier-Stokes方程在R^n中的前向自相似解
Forward self-similar solutions of the Navier-Stokes equations in R^n
arXiv 2609.33316 · 2026-09-29AgentLoop：面向工具增强型LLM智能体的槽闭合执行循环的运行时控制
AgentLoop: Runtime Control of Slot-closed Execution Loops for Tool-augmented LLM Agents
arXiv 2609.33315 · 2026-09-29ZeroGAR：零样本图模型对抗鲁棒性基准测试
ZeroGAR: Benchmarking the Adversarial Robustness of Zero-Shot Graph Models
arXiv 2609.33314 · 2026-09-29稀疏多项式系统实点计算的矩方法
The Moment Method for Computing Real Points of Sparse Polynomial Systems
arXiv 2609.33313 · 2026-09-29当隐私将机器学习中介的决策移至设备端：拍卖中的信息与激励错位
When Privacy Moves ML-Mediated Decisions On Device: Information and Incentive Misalignment in Auctions
arXiv 2609.33312 · 2026-09-29SocialHumanoid：通过一步式共语动作生成实现富有表现力的人形机器人行为
SocialHumanoid: Towards Expressive Humanoid Behavior via One-Step Co-Speech Motion Generation
arXiv 2609.33311 · 2026-09-29CompliantWBC：通过力潜在估计和残差阻抗目标实现重型人形机器人的全身柔顺控制
CompliantWBC: Whole-Body Compliance for Heavy Humanoids via Force Latent Estimation and Residual Impedance Targets
arXiv 2609.33310 · 2026-09-29跨生物学尺度的突变风险的机制解释
A mechanistic interpretation of mutation risk across biological scales
arXiv 2609.33307 · 2026-09-29LoopTrack：参数高效Transformer跟踪的简单基线
LoopTrack: A Simple Baseline for Parameter-Efficient Transformer Tracking
arXiv 2609.33306 · 2026-09-29一种用于BD-RIS辅助多用户波束成形的低维黎曼交替优化算法
A Low-Dimensional Riemannian Alternating Optimization Algorithm for BD-RIS-Assisted Multiuser Beamforming
arXiv 2609.33305 · 2026-09-29相关性并不意味着适用性：面向个人GUI代理的经验激活
Relevance Does Not Imply Applicability: Experience Activation for Personal GUI Agents
arXiv 2609.33304 · 2026-09-29BITS：重新思考不规则时间序列预测的公平与全面评估
BITS: Rethinking Fair and Comprehensive Evaluation for Irregular Time Series Forecasting
arXiv 2609.33303 · 2026-09-29钛的α相和ω相的低温热膨胀
Low-temperature thermal expansion of the $α$ and $ω$ phases of titanium
arXiv 2609.33302 · 2026-09-29犹豫感知的扩散语言模型在线策略蒸馏
Hesitation-Aware On-Policy Distillation for Diffusion Language Models
arXiv 2609.33301 · 2026-09-29Skoda $L^2$除法定理的一个逆定理 II
A Converse to the Skoda $L^2$ division theorem II
arXiv 2609.33300 · 2026-09-29直接隐藏状态对齐：映射与控制LLM中的偏好表达
Direct Hidden-State Alignment: Mapping and Controlling Preference Expression in LLMs
arXiv 2609.33298 · 2026-09-29你看到的错误并非你犯下的错误：面向推理错误定位的进展感知推理起源
The Error You See Is Not the Error You Made: Progression-aware Reasoning Origin for Reasoning Error Localization
arXiv 2609.33297 · 2026-09-29BaatCheet：面向印度语言对话翻译的多语种语料库
BaatCheet: A Multilingual Corpus for Dialogue Translation in Indian Languages
arXiv 2609.33296 · 2026-09-29TraceDance：一种从真实世界智能体部署轨迹构建智能体行为基准的自动化系统
TraceDance: An Automated System for Building Agent Behavior Benchmarks from Real-World Agent Deployment Traces
arXiv 2609.33295 · 2026-09-29高阶结构光定制多极光镊捕获
Tailored multipolar optical trapping with higher-order structured light
arXiv 2609.33294 · 2026-09-29首次阿尔茨海默病神经影像检查的时机：遗传、认知与社会因素的多中心分析
Timing of First Alzheimer's Neuroimaging: A Multicenter Analysis of Genetic, Cognitive, and Social Factors
arXiv 2609.33293 · 2026-09-29关于强制与反强制最小割的复杂性
On the Complexity of Forcing and Anti-Forcing Minimum Cuts
arXiv 2609.33292 · 2026-09-29多次响应的统计优势：从示范中学习的视角
The Statistical Benefits of Multiple Responses for Learning from Demonstrations
arXiv 2609.33291 · 2026-09-29校准，而非答案选择：在推理模型中蒸馏内部置信度
Calibration, Not Answer Selection: Distilling Internal Confidence in Reasoning Models
arXiv 2609.33290 · 2026-09-29
